Abstract:This paper provides a novel approach to understanding the nature of innovation and scientific progress by analyzing large-scale datasets of scientific literature. A new measure of novelty potential or disruptiveness for a set of scientific entities is proposed, based in the mathematical formalism of algebraic topology via a method called . In this framework, understanding where academic ideas depart from the existing body of knowledge to is key to scoring a set of entities and their potential for filling future knowledge gaps. This framework is motivated by the assumption that scientific discovery has underlying regularities that can be modeled and predicted. Our method uses a , which is a type of ranking of hypergraph components along a chosen parameter. In this work two different axes are used, which constructs a growing grid of sub-hypergraphs. The axes of time (scientific knowledge evolution) and normalized point-wise mutual information (network structure) affords the ability to succinctly represent the entire dynamic structure of the scientific literature network. We then find that using very simple and interpretable measures of centrality derived from this crude or affords the ability to predict links within the dynamic scientific network. While several different methods of link prediction have been proposed in the past, the method presented here the notion of link prediction to a higher dimension, as the boundary of the knowledge gap may be more than just 0-dimensional nodes. The system presented here not only suggests a mathematical basis, consistent with observations in cognitive neurosciences regarding early childhood language acquisition, but additionally provides useful applications for the scientific community in predicting and ranking hypothesis for scientific discovery.

Link prediction for interdisciplinary collaboration via co-authorship network

Modeling Scholarly Collaboration and Temporal Dynamics in Citation Networks for Impact Prediction

Link Prediction for Statistical Collaboration Networks Incorporating Institutes and Research Interests

A Link Prediction-Based Method for Identifying Potential Cooperation Partners: A Case Study on Four Journals of Informetrics

Collaboration Prediction in Heterogeneous Academic Network with Dynamic Structure and Topic.

Link prediction in author collaboration network based on BP neural network

Link Prediction Based On Clustering Information In Scientific Coauthorship Networks

Predicting Scientific Success Based on Coauthorship Networks

Clustering-Based Link Prediction In Scientific Coauthorship Networks

Link prediction and feature relevance in knowledge networks: A machine learning approach

Asymmetry of social interactions and its role in link predictability: the case of coauthorship networks

The link prediction problem for social networks

Persistent homology centrality improves link prediction performance in Pubmed co-occurrence networks

Interdisciplinary patterns of a University: Investigating collaboration using co-publication network analysis

Cooperative Relationship Prediction Between Scholars in Heterogeneous Academic Network.

Dynamic Network Analytics for Recommending Scientific Collaborators

Temporal Link Prediction Using Graph Embedding Dynamics

Scientific Collaborator Recommendation in Heterogeneous Bibliographic Networks

Co-authorship Prediction Based on Temporal Graph Attention

Evolving Coauthorship Modeling and Prediction Via Time-Aware Paired Choice Analysis.

A Methodology for the Analysis of Collaboration Networks with Higher-Order Interactions